Covid-19 Testing & Lean

Artist impression of Corona virus
Artist impression of the corona virus

Over a Zoom coffee recently, a friend had been telling me about her day at work, which hadn’t been great. She works in a doctor’s surgery and had received numerous phone calls from people waiting for their Covid-19 test results. They had been told they would get their results in 2 days but the tests were taking longer. The patients had to self-isolate while the tests were being processed, which was at the very least frustrating and in many cases impacting on their work or health. It seemed that her job would be a lot less stressful and more productive if the tests could be sped up so she’d get fewer phone calls chasing test results.

Go easy on the person, hard on the process

So, what has this got to do with lean? Many think lean is just something to do with manufacturing, but the principles apply to any process. And getting Covid-19 tests done is a process.

It’s no-one’s fault that the tests were taking longer that the promised 2 days. Certainly not the fault of a receptionist on the other end of the phone. Or the nurse taking the test. Or the lab technician processing the test. It’s the process, or system that needs fixing.

What is a process? It’s just a series of steps between a supplier and customer. In the case of Covid-19 testing, the patient is both the supplier and the customer. The steps would be something like this:

  1. Patient’s swab is taken at clinic
  2. Enter test details into the computer
  3. Label, bag sample and put in out-tray
  4. Sample sits in out-tray until picked up by courier
  5. Courier sample over to lab
  6. Sample waits for machine to be free
  7. Process the sample through the test machine
  8. Enter test results into computer
  9. Provide test results to patient                  

Disclaimer – this is for illustrative purposes only to show how lean principles apply. And if this were for real, the people in the process, who know the details, would be the ones analysing it, not me. And each step would be further broken down and the time taken for each step noted. I’m just using Covid-19 testing as an example of a (non-manufacturing) process.

Bottlenecks & WIP

Any process will only be as fast as its slowest step – the constraint or bottleneck. Usually increasing the capacity of any step other than the bottleneck will not increase the overall capacity. It’s easy to tell which step is the bottleneck as the work-in-progress (WIP) piles up before it. There is no flow.

In the case of Covid-19 testing, many drive-through and other clinics opened up quickly and people were encouraged to get tested. If the capacity of taking swabs exceeds the swab processing capacity, then there was going to be a problem, which would only get worse over time as the swabs (WIP) build up.

The more tests you have backed up in front of the bottleneck, the worse things get. You may get queue-jumping for some tests which are seen to be more urgent. And as anyone who has been in a queue knows, that means that others in the queue will take longer.

In an ideal system, there is no queue. In practice there may be buffers between steps, but the amount is planned for – not just a pile of tests waiting.

I imagine that test machines are expensive and it is difficult to just build more in a hurry. And train more people to use them. But there are other ways to increase the capacity of a machine. You need to measure its uptime (also called OEE – overall equipment effectiveness). Is it running at 100%, no quality issues, no breakdowns? The time between tests, time setting up the machine, putting information into the system, checking names or other necessary paperwork are all opportunities to increase capacity quickly.

Pull vs Push

If testing capacity is increased through more drive-through and other clinics, these test swabs are ‘pushed’ onto the next step – processing the tests. Tests at clinics may be done as fast as possible, particularly if there is a target number to reach each day. Having a large pile of tests to process doesn’t make the processing any faster. It can even slow things down as there are more tests to sort through and possibly prioritise if some urgent ones come along. Or decide which to do first. Or be stressed by the big pile knowing everyone is relying on you to speed up, which is when mistakes can happen.  How do you ensure that there is ‘first in first out’ with a big unmanageable pile of tests to process? Test times will probably vary, so some customers will get their results sooner than others.

What can work better is using ‘pull’. You need to know the customer demand rate (how many tests do we need to get back to customers) and work back through the system at this same rate. Any constraints or bottlenecks within the system must operate at the required customer rate. The capacity of the bottleneck steps may need to be increased, but until you do that, it’s important to be honest with your customers. Only promise to deliver what you can achieve, which is the rate of tests through the bottleneck. Other steps in the process, which have more capacity than customer demand don’t have to run flat out. This bit is often hard to get your head around, but to optimise a process, some steps will slow down and that’s OK! It’s only the bottleneck that has to run flat out.  

Value vs Waste

Lean is a very simple concept. It looks for waste and reduces it. Waste is anything the customer doesn’t value. In the example of being tested for Covid-19, the only steps that add value are those in bold above – getting the swab taken, processing it through the testing machine and getting told about the result. If it takes 10 minutes to take a swab, 3 hours to process a test and 10 minutes to send a text result, then the value-add time is <4 hours. Even with a promised turnaround of 2 days, value is only being added 4/48 or 8% of the time. 92% of the time is waste such as waiting in a queue, waiting for the courier pick up time or being couriered to the lab. The longer the queue, the more waste and WIP and the longer it will take for someone to get their test results back.

So whether you’re involved in Covid-19 testing, manufacturing or any other business operation, you have processes. Understanding the steps in a process, the capacity of its bottleneck and applying lean principles can improve processes usually without spending a lot of money. There are benefits too as the customers’ experiences improve, which in turn leads to employees’ jobs being less frustrating and stressful. In the case of Covid-19 testing, there is no reason why a 2 day turnaround (or better) can’t be guaranteed as long as too many tests aren’t pushed into the system.

Let’s hope those involved with Covid-19 testing have a good understanding of the bottlenecks of their processes and have improved things, so my friend’s day at work will also improve. I’ll find out at our next coffee catch up, which happily will be in person, rather than via Zoom.

Lean, Six Sigma and Theory of Constraints

Process improvements. Which should I use? Lean, Six Sigma or TOC?

There is an ongoing debate between consultants on which is better – lean, six sigma or TOC (theory of constraints). There is a tendency to use baffling jargon – often in japanese – and overcomplicate things. In reality most continuous improvement tools are similar, just with different focuses:

  • Lean focuses on minimising waste.
  • Six Sigma focuses on minimising variation.
  • Theory of Constraints focuses on reducing bottlenecks.

An approach that usually works is to use TOC to help you decide which part your process to tackle first. You can then use some lean or six sigma tools release capacity.

Inventory is what we use to cope with less than perfect processes as it hides issues such as delays, poor quality, process time variations, supply shortages etc. Lean and six sigma will tend to push to reduce all inventory, while TOC will accept that some inventory is of value as a buffer. This is especially true for jobbing shops, where day to day balancing of work is difficult.